41　Verification of Test Model Data for Failure Detection Using Logistic Regression Detectable String Failures Including Cluster Failures

Abstract

In recent years, with the spread of photovoltaic systems (PVSs) are becoming deterioration and failures that lead to accidents and fires are increasing, and reducing the cost of maintenance and inspections to prevent these problems is an issue. In this study, we propose a machine learning method using logistic regression analysis, which is a type of multivariate analysis and has a low computational load, with the aim of simplifying maintenance and inspection. In this presentation, we will report that we have verified the effectiveness of the proposed method by learning using measured data from a site of PVS.